DQN Agent

Deep Q-Network agent implemented in Python capable of learning to land on the moon in MoonLander-v2 environment from AI Gym library.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

• Pycharm 2019.x
• Python 3.6
• Tensorflow 1.13
• Keras 2.2.4
• Jupyter Notebook 5.7.8

Installing

1. Install Python and Pycharm
2. Clone this repository to your local drive
3. Open root directory in Pycharm and let it fetch third-party dependencies from requirements.txt
4. Try to run dqn_example.py. In case of any failures add above versions of Tensorflow, Keras and Jupyter Notebook in the requirements.txt

Running the tests

1. Open report.ipynb with Jupyter Notebook and try executing code lines

Deployment

1. After applying changes to any python script, update the content of report.ipynb

Built With

• PIP - Python package installer

Authors

• Jakub Gros
• Sebastian Kucharzyk

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments

• Code style guide used: PEP 8 (https://www.python.org/dev/peps/pep-0008/)